Replicating Proxmox VE virtual machines (VMs) to a remote site is a common requirement for disaster recovery (DR) and business continuity. While Proxmox does not have native multi-site replication in the same way as VMware SRM or Hyper-V Replica, you can still achieve efficient remote replication using built-in tools and standard Linux techniques.
1. Proxmox VM Replication (Built-in, but for same-cluster only)
Proxmox supports VM replication natively between nodes in the same cluster, but this doesn’t work across WAN-separated locations unless you stretch the cluster (not recommended over high-latency links).
Great for: Local datacenter failover
Not designed for true remote/offsite DR
2. Using Proxmox Backup Server (PBS) + Remote Sync
Proxmox Backup Server (PBS) is the recommended way to replicate VM data to a remote site securely and efficiently.
How it works:
- Install PBS at both primary and remote sites.
- Set up a datastore sync job from the primary PBS to the remote PBS.
- Backup VMs/CTs at the source using Proxmox VE → PBS.
- PBS automatically sends incremental backups to the remote site.
Features:
- Deduplication and compression
- Encryption at rest and in transit
- Incremental sync = efficient over WAN
- Supports fast full restores and file-level recovery
Steps:
# On remote PBS, create a new datastore
# On primary PBS, add a sync job pointing to the remote PBS
proxmox-backup-manager sync-job create \
--source remote-pbs.example.com:8007 \
--store remote-store-name \
--remote remote-site-name \
--schedule "daily"
Combine with a Proxmox VE host at the remote site to restore VMs on demand if disaster occurs.
3. ZFS Send/Receive Replication
If you’re using ZFS storage for your VMs:
How it works:
- Create ZFS snapshots of VM disks
- Use
zfs send | ssh zfs receive
to stream snapshots to the remote Proxmox server
Example:
# On source node
zfs snapshot rpool/data/vm-100-disk-0@daily
# Send to remote
zfs send -R rpool/data/vm-100-disk-0@daily | ssh root@remote-site zfs receive -F rpool/backup
Pros:
- Fast, efficient block-level transfer
- Works over WAN with SSH
- Can automate with cron or replication tools like syncoid
Cons:
- Manual VM config sync required (
/etc/pve/qemu-server/100.conf
) - Some expertise needed to manage reliably
4. Backup to Cloud or Remote Object Storage (Optional)
If your remote site includes an S3-compatible store (like MinIO, Wasabi, or TrueNAS with S3), PBS can backup directly to object storage as a long-term DR option.
Choosing the Right Replication Strategy
Method | Best For | Pros | Cons |
---|---|---|---|
Proxmox Cluster Replication | Same-site, <10ms latency | Built-in, fast | Not for remote sites |
PBS + Sync Jobs | Remote DR, offsite backups | Secure, efficient, incremental | Requires PBS at both sites |
ZFS Send/Receive | ZFS-only deployments | Fast, scriptable, WAN-capable | Manual restore setup, no GUI |
Veeam/3rd party backups | Multi-platform DR | VM/file-level restore, scheduling | External software + agents required |
Final Notes
- Proxmox Backup Server (PBS) is the most robust and production-friendly solution for offsite replication.
- ZFS send/receive is ideal for advanced users or lightweight setups.
- Stretching a Proxmox cluster across WANs is not recommended due to latency and quorum issues.
- Consider using DNS failover or VPN tunnels for secure replication traffic.
Need Help Designing a DR Plan?
We can help with:
- Multi-site Proxmox + PBS setup
- ZFS replication automation
- VM failover and DR runbooks
- Integration with cloud storage or Veeam
Get in touch with Saturn ME today for a free Proxmox consulting session—no strings attached.